numexpr

The numexpr package evaluates multiple-operator array expressions many times faster than NumPy can. It accepts the expression as a string, analyzes it, rewrites it more efficiently, and compiles it on the fly into code for its internal virtual machine (VM). Due to its integrated just-in-time (JIT) compiler, it does not require a compiler at runtime.

homepage: http://code.google.com/p/numexpr/

version	versionsuffix	toolchain
2.5.2	-Python-2.7.11	intel/2016a
2.6.1	-Python-2.7.12	foss/2016b
2.6.1	-Python-2.7.12	intel/2016b
2.6.1	-Python-3.5.2	intel/2016b
2.6.4	-Python-3.5.1	foss/2016a
2.6.4	-Python-2.7.13	foss/2017a
2.6.4	-Python-3.6.4	foss/2018a
2.6.4	-Python-3.6.1	intel/2017a
2.6.4	-Python-3.6.3	intel/2017b
2.6.4	-Python-2.7.14	intel/2018a
2.6.4	-Python-3.6.4	intel/2018a
2.6.5	-Python-2.7.15	foss/2018b
2.6.5	-Python-3.6.6	foss/2018b
2.6.5	-Python-3.6.6	fosscuda/2018b
2.6.5	-Python-2.7.15	intel/2018b
2.6.5	-Python-3.6.6	intel/2018b
2.7.0		intel/2019a
2.7.1	-Python-3.7.4	foss/2019b
2.7.1	-Python-3.8.2	foss/2020a
2.7.1	-Python-3.7.4	fosscuda/2019b
2.7.1	-Python-2.7.16	intel/2019b
2.7.1	-Python-3.8.2	intel/2020a
2.8.1		foss/2021a
2.8.1		intel/2021a
2.8.4		foss/2022a
2.9.0		foss/2023a

(quick links: (all) - 0 - a - b - c - d - e - f - g - h - i - j - k - l - m - n - o - p - q - r - s - t - u - v - w - x - y - z)